The mixed traffic flow has complex dynamics by nature. The kinematic differences between automobiles and motorcycles result to distinct driving behaviors. Traditional automobile-based traffic flow theory is not always suitable for mixed traffic streams. The purpose of this study is to observe from actual data a clearance boundary, called Safety Space, drivers maintain from other vehicles, and use it as a spatial filter to determine conflicts in mixed traffic flows. Image data are collected from an Unmanned Aerial Vehicle (UAV), and microscopic characteristics such as vehicle type, position, velocity, and trajectory are extracted through computer vision techniques. The Histogram of Oriented Gradients (HOG) feature and the Support Vector Machine (SVM) classifier are utilized for the vehicle detection, while the Kalman Filter is employed for the derivation of vehicle trajectories. The Safety Space is then determined based on those trajectories. Validation data are collected at intersections in Taipei, Taiwan; Bangkok, Thailand; and Mumbai, India. The vehicle detection and tracking are satisfactory, and the Safety Space surrogate reveals risk zones caused by spatial proximity between vehicles.

Conflict analytics through the vehicle safety space in mixed traffic flows using UAV image sequences

This paper proposes an overall Multi-Controller Architecture (MCA) for safe and flexible navigation of autonomous navigation, under uncertainties in highway use-cases. In addition to the details given about the main modules (and their interactions) composing the proposed MCA, an important focus of the paper is made on the definition of a robust Two-Sequential Level Decision Network (TSLDN), which uses both: Extended Time-To-Collision (ETTC) metric and a new definition of a specific Predicted Inter-Distance Profile (PIDP, between vehicles during lane changes maneuvers) in order to estimate the maneuvers risks. The TSLDN is utilized for: the driving situation assessment, decision-making and for safety retrospection over the current maneuver risk. It allows us to have the best decision to achieve the vehicle navigation task while maximizing its safety. Several simulation results show the good performance of the overall proposed control architecture, mainly in terms of efficiency to handle probabilistic decision-making even for very risky scenarios.

Multi-Level Bayesian Decision-Making for Safe and Flexible Autonomous Navigation in Highway Environment

This paper describes the design of an overall Multi-Controller Architecture (MCA) for safe automated driving, under uncertainties in highway environment. This MCA combines the decision-making process, the path planning and the control algorithms. In order to ensure the safety and smoothness of the vehicle navigation, a decision-making strategy for handling lane change maneuvers is proposed by means of a robust Two-Sequential Level Bayesian Decision Network (TSLDN). The latter is utilized for the driving situation assessment, the decision-making and for safety verification of the current performed maneuver. Moreover, from actual navigation risk assessment standpoint, a dual-safety criterion combining an Extended Time-To-Collision (ETTC) and a novel Dynamic Predicted Inter-Distance Profile (DPIDP) is developed. The DPIDP is evaluated on-line as the actual distance between vehicles during lane changes maneuvers over an observed control horizon. Thanks to this proposed algorithm, a safety retrospection over the current maneuver risk could be carried out. In this way, the overall MCA described in this paper allows the best probabilistic decision to achieve the vehicle navigation task in hazardous situations while maximizing its safety. Several simulation results show the good performance of the overall proposed control architecture, mainly in terms of efficiency to handle probabilistic decision-making even for very risky/complex scenarios.

Safe Autonomous Overtaking Maneuver based on Inter-Vehicular Distance Prediction and Multi-Level Bayesian Decision-Making

Risk Management for Intelligent Vehicles based on Interval Analysis of TTC

A system for efficiently calculating the risk of the mobile system colliding with a static obstacle, estimating the resulting time to collision and augmenting the user-commanded velocity commands is presented in this paper. The innovation of the approach is in merger of the laser range finder sensor measurements with the motion model of the mobile system and predicting the virtual path of the obstacles in the local coordinate system of the sensor. The resulting trajectories are then geometrically evaluated and intersections with the line-based mobile system outline are detected - approach avoids discretization of either temporal or spatial prediction horizons and produces the time-to-collision estimates with an analytical method. This not only improves the accuracy of the results, but also reduces the computational and storage loads of the processing system. The paper also presents a collision prevention system for a mobile system that is controlled by the human operator, allowing safe operation either for remote operation or use by the untrained/unreliable operator. The system was implemented on a rehabilitation mobile platform and preliminary results are provided and evaluated. In summary, the results show that the method is promising and easily applicable to target mobile systems.

Efficient Time-To-Collision Estimation for a Braking Supervision System with LIDAR

Ways to estimate the time-to-collision are explored. In the context of traffic simulation models, classical lane-based notions of vehicle location are relaxed and new, fast, and efficient algorithms are examined. With trajectory conflicts being the main focus, computational procedures are explored which use a two-dimensional coordinate system to track the vehicle trajectories and assess conflicts. Vector-based kinematic variables are used to support the calculations. Algorithms based on boxes, circles, and ellipses are considered. Their performance is evaluated in the context of computational complexity and solution time. Results from these analyses suggest promise for effective and efficient analyses. A combined computation process is found to be very effective.

/pdf/new-algorithms-for-computing-the-time-to-collision-in-36tewa332a.pdf

Jia Hou

Papers

New algorithms for computing the time-to-collision in freeway traffic simulation models